Political polls are just noise that's best ignored, by John McGauley

I read the other day that a poll showed Joe Biden ahead of Donald Trump in the Lone Star State. Are you kidding me? I’ll walk down Main Street in Keene dressed as Ginger from “Gilligan’s Island” if a Democratic presidential candidate carries Texas in 2020.

Polls aren’t used anymore to accurately reflect voter preferences, but instead to manufacture little newsburgers that can sustain a headline that readers, viewers and listeners will repeat to friends. “I heard Joe Biden is beating Trump in Texas,” is the desired result.

You can get a poll to produce anything you want by manipulating the questions or a dozen other parameters of the survey.

The august Pew Research Center puts it this way:

“Determining voter preference among the candidates running for office would appear to be a relatively simple task: just ask them who they are going to vote for on Election Day. In fact, differences in how this question is asked and where it is placed in the questionnaire can affect the results.”

Talk about understatement. Polls are partisan spears.

Here’s another thing. Who has landlines anymore, which pollsters have traditionally used to poll Americans? And who answers spam calls on their cellphones? Who responds to surveys that pop up on your PC or smartphone?

So, who responds to polls? My suspicion is they are primarily zealots on both sides of the political spectrum, hardly a sampling of Americans.

It gets worse. Many polls these days are what the polling industry calls IVRs, technically called Custom Interactive Voice Response Telephone Polling Solutions, which integrate automatic phone systems and software that call individuals and conduct telephone surveys. Another form of an IVR survey is called a computer-assisted telephone interview (CATI) application.

Put more simply, these are robocalls, using the same technology as those phony calls claiming to be the IRS warning you that they’re just about to garnish your bank accounts unless you call their toll-free number right away and tell them your Social Security number.

So, unless pollsters are using some new technique that’s not been revealed, I don’t know how they can claim they’re legitimate.

I guess you could walk down Main Street in Keene and ask every third person what they think, but you’d only get a tiny sample from a New England village. To draw any inferences, you’d have to do the same in Enid, Okla., Augusta, Ga., and Bend, Ore. But even if you did this, you’d miss the people who don’t or can’t walk downtown, don’t eat in downtown restaurants, don’t work downtown or who instead shop online. No survey can claim to be random or represent everybody.

People screw with pollsters, too, giving answers with the snarky intent to mess things up. Or, people answer questions with responses they feel are more politically correct, rather than what they actually believe. That is, if they even bother to answer.

Who pays for the polling? The answer can be murky. Often it’s a political action committee that’s a proxy for a specific candidate, cause or party.

What was the size, scope and makeup of the sample of people polled? When was the poll conducted? What was the “response rate” in the poll, and by what conveyance was it conducted? There’s no way those questions are answered in a simplistic seven-second TV news bite or a three-paragraph story.

After the 2016 election, the online numbers-crunching site FiveThirtyEight looked at how the top 20 polling organizations in the United States performed. It showed all the pollsters were wrong in predicting the Trump victory, including FiveThirtyEight itself. All except one of the 20 showed Hillary Clinton winning and the two worst-performing of all the polls were Google Surveys and our very own University of New Hampshire poll, which overestimated Democrats’ performance by an average of almost 9 percentage points in the polls it conducted of New Hampshire and Maine.

I think those polls got it wrong in 2016 because potential Trump voters kept very quiet about their preferences until they actually were in the voting booth.

Besides, until the Democrats nominate a candidate, polls are especially meaningless. How can a poll determine how Trump measures up against a specific Democrat?

There are very few independents when it comes to Donald Trump. Voters’ views of the man are baked in the cake and have been since shortly after he was elected. Americans either hate him or love him. I doubt anyone is ambivalent about the man.

I’ve said this a score of times before: Donald Trump is not a politician, and has a temperament and background unlike any president in our long history. Every president we’ve elected before Trump was a professional politician who’d run for numerous elected offices prior to landing in the White House, or had been a decorated military leader. Trump has no military background and never ran for as much as dog catcher before he landed on Pennsylvania Avenue.

You’re going to hear “news” of hundreds of polls in the coming months, and should ignore them all; their intent is not to inform but to raise the clanging of cymbals.

The old political bromide is there is only one poll that matters, and that’s the one conducted on Election Day. The result will hinge on whether more of the Trump haters or lovers turn out to vote.

My prediction is that 2020’s election will closely parallel the last presidential race, coming down to an Electoral College battle. Fasten your seatbelts, ignore the polls and don’t listen to your friends, as you tend to hang out with like-minded people who cannot legitimately reflect the sentiments of 138 million Americans, which is the number who voted in 2016. Oh, and by the way, that means 60 million citizens eligible to vote in 2016 never showed up at the polls.

John McGauley, an author and local radio talk-show host, writes from Keene. He can be contacted at mcgauleyink@gmail.com

Latest e-Edition